A Distributed Whole Genome Sequencing Benchmark Study

Area of Science:

• Genomics
• Bioinformatics
• Data Science

Background:

• Population sequencing initiatives involve multiple distributed centers, necessitating consistent data quality.
• Ensuring accuracy across different sequencing sites is crucial for large-scale genomic studies.

Purpose of the Study:

• To assess the impact of sequencing centers and analysis pipelines on whole genome sequencing data quality.
• To evaluate the reproducibility and concordance of genomic data generated across distributed sites.

Main Methods:

• Whole genome sequencing of DNA replicates from cell lines across three Canadian centers.
• Application of site-specific secondary analysis pipelines to generated sequence data.
• Quality assessment using concordance with benchmark variant truth sets and three-way concordance analysis.

Main Results:

• Variant concordance between datasets was similar across different sequencing centers and replicates when using the same analysis pipeline.
• Statistically significant differences in datasets were primarily attributed to the analysis pipeline, not the sequencing center.
• Standardized and updated analysis pipelines can unify data processing across sites.

Conclusions:

• Genome sequencing projects can achieve reliable quality and reproducibility across a network of distributed sites.
• The choice of analysis pipeline significantly impacts data consistency, emphasizing the need for standardization.
• Collaborative sequencing efforts can be confident in the aggregate data quality generated network-wide.